358 WRITINGS OF JOSEPH HENRY. [1855 



was diminished. A corresponding diminution was also pro- 

 duced in the amount of radiant light given off, and this was 

 readily perceptible to the sight. This effect was not due to 

 the cooling of the flame by the conduction of the mica, since 

 it is almost a non-conductor of heat, and this property was 

 exhibited by the fact that the luminosity of the mica was 

 confined to that part which was at the surface of the flame 

 on either side. 



It appears therefore from these experiments, that the intro- 

 duction of a solid of great radiating power into a mass of 

 materials in a state of combustion, increases the amount of 

 heat thrown into the space around, without increasing the 

 absolute quantity produced by combustion, the increase of 

 radiant heat being at the expense of the heat of combination. 

 To give a practical illustration of the condition of the matter, 

 if a given quantity of fuel is employed in evaporating water, 

 by combustion under a kettle, the useful effect would be 

 diminished by inserting in the flame beneath or amid the 

 combustible a better radiating substance than itself, while in 

 the case of a fire to warm a room the effect would be directly 

 opposite; a greater amount of heat would be thrown into 

 room, and less of the heat of combustion would be carried 

 up the chimney with the escaping gas. Or to give another 

 example. If over a coal fire a boiling pot be suspended, and 

 a roasting oven before it, the introduction of a radiating 

 material would increase the effect on the latter at the expense 

 of that on the former. 



Count Rumford has elsewhere shown that flame is a bad 

 conductor of heat, and in stoves and boilers heated by flame it 

 is therefore necessary that the draft be made to impinge with 

 considerable force upon projecting portions of the metal in 

 order that the greatest amount of heat may be absorbed. 



If a column of heated air moves rapidly through a per- 

 pendicular stove-pipe, but a comparativelj'^ small portion of 

 the heat will be absorbed by the metal and radiated into 

 space around. A cylindrical stratum of non-conducting air 

 in contact with the metal will be comparatively at rest, and 

 through this the moving column of heated air will rapidly 



